Reply to Bovet et al.

Bovet et al. provide useful additional information that is certainly relevant to the broader context of our article 1. As our review and our search terms focused on the more commonly used socioeconomic status (SES) indicators (income/wealth, education, occupation or composite), our search criteria did not capture the paper mentioned by Bovet et al. 2, which used type of school (public vs. private) as SES indicator. Despite our focus on the more common SES indicators, our search strategy identified one study on children that did use school type as SES indicator 3, because the abstract and title of this study did include terms that matched our search terms, while the Bovet et al. study did not.

School type may well become a more widely accepted and validated SES proxy in the future and should then be included in future systematic reviews, but at present, more research is needed to examine the validity of the measure. An encouraging example of one such validation, if in a different context, is by Piovesan et al. 4.

Nevertheless, the findings of the Bovet et al. paper 2 still support our conclusion that – based on the still very limited amount of evidence – obesity/overweight among children in low and middle income countries appears to remain more prevalent in higher SES groups (proxied by private school attendance in 2) compared with those in lower SES groups (proxied by public school attendance). We also note that even though only one study in Africa fulfilled our search/selection criteria on children, our review did include six studies undertaken in Africa (including one in Seychelles) on adults.

Other studies presented in the letter were published either prior to the period we covered 5 or after 6, 7. We did not review 8 because it was undertaken in developed countries, although we have cited it (see ref#50 in 1), where we found appropriate.